Fire safety engineering has become increasingly central to construction as urbanisation drives larger projects and hybrid building methods proliferate. Complex developments that mix multiple functions and occupancy types demand particularly rigorous fire safety design to ensure structures can withstand fire scenarios while enabling safe evacuation and limiting damage.

Sitowise Ltd, a Nordic technical consulting and digital solutions firm operating primarily in Finland and Sweden with roughly 1,700 employees, counts fire safety engineering among its core services. The company's structural fire engineering work ensures buildings perform safely under both routine and emergency conditions, relying on extensive computational analysis to model fire spread, smoke behaviour, and temperature effects on structures and evacuation routes.

Simulation demands outpaced internal computing capacity

Valtteri Korventausta, Head of Department at Sitowise's Fire Engineering Unit, explains the computational challenge: "In fire engineering, we use simulations to understand, for example, how a fire spreads, how smoke behaves, and how temperatures affect structures and evacuation conditions. Simulation requires large amounts of data, which in turn requires significant computing power."

As Sitowise took on more demanding projects—including shopping centres, sports arenas, large industrial facilities, and tunnel and underground construction—computational requirements grew, slowing the design process. The company began investigating alternatives to upgrading its own hardware.

Discovery through peer recommendation

A colleague working on a joint project introduced Korventausta to the possibility of accessing high-performance computing through LUMI. "We were considering upgrading our own computing hardware when a colleague in fire engineering told us about the possibility of using the LUMI supercomputer," Korventausta recalls.

He reached out to CSC, which offered access through the Try&Buy programme—a free evaluation initiative allowing companies to test whether LUMI suits their needs before committing resources. According to Jarno J. Laitinen, CSC's Account Manager for the Sitowise relationship, "The Try&Buy project includes user accounts for two users, CPU and GPU resources, data storage capacity for testing purposes, and most importantly, free expert support for implementation."

The typical evaluation window runs three to six months, after which organisations can either transition to a paid computing services agreement or discontinue use. Despite neither Korventausta nor his team having prior high-performance computing experience, adoption proved swift and smooth. "Because the colleague who recommended LUMI was using the same software we use, the necessary tools were already available on LUMI, and expert support was readily accessible, allowing us to get started quickly," Korventausta notes.

Tangible gains in speed and precision

LUMI has delivered measurable improvements to Sitowise's fire engineering capabilities. The supercomputer enables substantially greater computational capacity, permitting simultaneous simulations and finer mesh resolutions that boost accuracy. "The biggest benefit is speed. Previously, simulations could take more than a week. Now we get results in just a few days," Korventausta explains.

Within the same project timeline, teams can now evaluate multiple design alternatives rather than committing to a single approach. This flexibility supports earlier error detection, smoother project management, and faster client deliverables. "When simulations are completed more quickly, the entire project progresses more smoothly. We can deliver results faster and move projects forward without unnecessary delays."

Accelerated computation also enables design optimisation based on actual temperature conditions, reducing unnecessary overdesign and refining material use. "Simulation provides input data that allows us to reduce structural weight without compromising safety."

Sitowise views LUMI as essential infrastructure for future engineering work. "As long as computing power is available this cost-effectively through LUMI, we see no need to invest in our own hardware. It gives us flexibility and allows us to focus on our core design work."

Environmental alignment and future capacity

Sitowise's environmental strategy targets greenhouse gas reductions across its operations and value chain while advancing sustainable construction practices. LUMI supports these goals: the supercomputer and its Kajaani data centre rank among the world's most environmentally efficient supercomputing facilities, with waste heat recovered for district heating and energy-efficient operations that reduce taxpayer costs.

CSC's Laitinen emphasises LUMI's role as a benchmark for Europe's ICT sector in meeting EU climate and green transition objectives. "It is a pleasure to work with companies that share similar environmental values. I would also like to remind people that LUMI is used exclusively for research, development, and innovation activities. It is used, among other things, for climate change modeling, medical research, and the development of European AI applications."

Europe's supercomputing infrastructure will expand further. LUMI-AI, scheduled for launch at the end of 2027, will nearly double current high-performance computing capacity and increase artificial intelligence capacity tenfold. The LUMI AI Factory will also house LUMI-IQ, a 150-qubit quantum computer ordered from IQM, within CSC's Kajaani data centre.